tgutil package - Tanay’s group collection of simple utility functions

Installation

You can install tgutil with:

remotes::install_github("tanaylab/tgutil")

Read legacy csv’s

fread_rownames

fread_rownames(..., row.var = "rowname")

Use fread() to read a csv/tsv with row names (e.g. one created with read.table()). The name of the column that will hold the row names is given in row.var.

All arguments are passed to the underlying fread() call. Note that the following arguments can’t be used (and will be silently ignored): header, skip, col.names.

Sparse Matrix I/O

fwrite_mm

fwrite_mm(x, fname, sep = " ", row.names = TRUE, col.names = TRUE)

Write the contents of the sparse matrix x into the text file fname in the MatrixMarket format (see http://math.nist.gov/MatrixMarket/formats.html).

The MatrixMarket format does not save row and column names. fwrite_mm() allows saving them into separate files using the row.names and col.names arguments. These arguments may contain file names into which the column and row names will be written. Alternately setting these arguments to TRUE will cause the row and column names to be written to <fname>.rownames and <fname>.colnames respectively.

The argument sep indicates the field separator that will be used. Note that using any field separator other than space will result in an output that is not compliant with the MatrixMarket format.

fread_mm

fread_mm(fname, sep = " ", row.names = TRUE, col.names = TRUE)

Reads a sparse matrix in the MatrixMarket format from the file fname.

The MatrixMarket format does not save row and column names. fread_mm() allows reading them from separate files using the row.names and col.names arguments. These arguments may contain file names from which the column and row names will be read. Alternately setting these arguments to TRUE will cause the row and column names to be read from <fname>.rownames and <fname>.colnames respectively.

Dense scatter plots

geom_dense_scatter

library(tgutil)
library(ggplot2)
n <- 1e4

# Create large dataset with non-linear relationship
x <- runif(n, -3, 3)
df2 <- data.frame(
    x = x,
    y = sin(x) * 2 + rnorm(n, 0, 0.5)
)

# Visualize non-linear relationship with density
ggplot(df2, aes(x, y)) +
    geom_dense_scatter(
        pal = c("gray90", "gray50", "orange", "red"),
        size = 0.4,
        alpha = 0.8
    ) +
    labs(title = "Non-linear Pattern with Density Coloring")